Junb Potentiates Function of BRCA1 Activation Domain 1 (AD1) Through a Coiled-Coil-Mediated Interaction

Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

JunB potentiates function of BRCA1 activation domain 1 (AD1) through a coiled-coil-mediated interaction

Yan-Fen Hu and Rong Li1 Department of Biochemistry and Molecular Genetics, School of Medicine, University of Virginia, Charlottesville, Virginia 22908, USA

BRCA1 is involved in the regulation of multiple nuclear events including transcription. AD1, one of the two trans-activation domains in BRCA1, stimulates transcription in a cell context-dependent manner. Here, it is shown that BRCA1 interacts with Jun proteins via a coiled-coil motif in AD1 and the basic leucine zipper (bZIP) region of the Jun proteins. The Jun-interacting domain in BRCA1 is critical for AD1-mediated transcriptional activation. In particular, the strength of AD1 in transcriptional activation is limited by the JunB level and ectopic expression of JunB potentiates the transcriptional activity of AD1. Furthermore, JunB mRNA expression is down-regulated in many ovarian tumor tissues examined. Thus, the coiled-coil-mediated cooperation between BRCA1 and JunB may facilitate the function of these proteins in tissue-specific transcriptional regulation and tumor suppression. [Key Words: BRCA1; AP1; Jun, transcription; bZIP; coiled-coil] Received March 26, 2002; revised version accepted May 2, 2002.

Mutations in BRCA1 account for a significant proportion ered (Hu et al. 2000; see Fig. 1A). Moreover, a highly of hereditary breast and ovarian cancers (Welcsh and conserved coiled-coil motif in AD1 is critical for its func- King 2001). Intense research in the past several years has tion in transcriptional activation. Interestingly, this implicated BRCA1 in regulation of multiple aspects of coiled-coil region is located immediately upstream of a nuclear function including transcriptional activation, demarcation point for cancer-predisposing mutations of DNA repair, recombination, and checkpoint control BRCA1 in which, according to a phenotype-genotype (Zhang et al. 1998b; Monteiro 2000; Scully and Livings- correlation study, change in ovarian cancer risks occurs ton 2000; Zheng et al. 2000; Parvin 2001). When tethered (Gayther et al. 1995). Therefore, it has been suggested to a transcriptional promoter, the BRCA1 carboxyl ter- that this region of the protein may contain a functional minus (BRCT) domain can stimulate transcription and domain that specifically protects ovarian epithelial cells remodel chromatin (Chapman and Verma 1996; Mon- from developing tumors (Rahman and Stratton 1998). teiro et al. 1996; Hu et al. 1999). The functional rel- In comparison with AD2, transcriptional activation by evance of these studies is underscored by the observation AD1 is cell-type dependent and less robust (Hu et al. that cancer-predisposing mutations in the same region 2000). In some cell lines (e.g., ES2, an ovarian cancer cell abolish the activity of the BRCT domain in transcription line), AD1 by itself exhibits very modest transcriptional and chromatin remodeling. Consistent with its potential activity, but it can synergistically stimulate transcrip- role in transcriptional regulation, BRCA1 is associated tion with AD2, whereas in other cell lines (e.g., with the RNA polymerase II holoenzyme and chroma- HEK293T, an embryonic kidney cell line), AD1 does not tin-modifying proteins (Scully et al. 1997; Neish et al. confer transcriptional stimulation, either alone or with 1998; Yarden and Brody 1999; Bochar et al. 2000; Pao et AD2. The molecular basis for the cell-context depen- al. 2000). In addition, the full-length BRCA1 protein can dence of AD1 activity remains to be elucidated, but one potentiate transcription from several natural promoters likely possibility is that a putative partner(s) that medi- (Somasundaram et al. 1997; Ouchi et al. 1998; Zhang et ates AD1 function may be limiting in these cells. al. 1998a; Harkin et al. 1999; MacLachlan et al. 2000). Herein, we find that AD1 interacts with the Jun pro- In addition to the BRCT domain (AD2), a second trans- teins of the AP1 family. We show that the BRCA1-Jun activation domain of BRCA1 (AD1) was recently discov- interaction is mediated by the coiled-coil region of BRCA1 and the bZIP region of the Jun proteins. Further- more, the cellular level of JunB is an important determi- 1Corresponding author. nant for the potency of AD1 in transcriptional activa- [email protected]; FAX (434) 924-5069. Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/ tion. We show that the mRNA level of JunB is down- gad.995502. regulated in the majority of ovarian tumor tissues

GENES & DEVELOPMENT 16:1509–1517 © 2002 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/02 $5.00; www.genesdev.org 1509 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Hu and Li

ducted a yeast two-hybrid screen using AD1 as the bait and a prey library of human ovary cDNA. To reduce the high background due to the intrinsic transcriptional activity of AD1, 100 mM 3-aminotriazole (3-AT) was included in the medium throughout the screen. A large- scale screen uncovered multiple independent cDNA clones that encode human JunB and JunD, two members of the activation protein 1 (AP1) family (Fig. 1A). The AP-1 family of transcription factors consists of homodimers and heterodimers of the Jun and Fos subfami- lies that play important roles in diverse aspects of cell proliferation and differentiation (Mechta-Grigoriou et al. 2001; Shaulian and Karin 2001). Mammalian Jun proteins include c-Jun, JunB, and JunD; Fos proteins include c-Fos, FosB, Fra1, and Fra2. A common characteristic of the AP1 family members is the presence of the basic leucine zipper (bZIP) motif that serves as the DNA-binding and dimerization domains (Chinenov and Kerppola 2001). The interactions between AD1 and the Jun proteins detected in the yeast system were specific, as JunB and JunD failed to bind to other known trans-activation domains, including AD2 of BRCA1, and the activation domains of several other mammalian transcription factors such as p53, Sp1, and CTF1 (Fig. 1A). Previous work has shown that a mutation at one of the key leucine residues in the coiled-coil motif in AD1 (L1407P) abolishes its function in transcriptional activation (Hu et al. 2000). The same mutation also abrogated the ability of AD1 to bind to the Jun proteins (Fig. 1A). Interestingly, all partial cDNA clones of JunB and JunD isolated from the screen encode the bZIP domain, suggesting that the AD1-Jun interaction may be mediated by the coiled-coil region of BRCA1 and the bZIP region of the Jun proteins. To confirm the yeast two-hybrid results, we also used a mammalian two-hybrid assay in which GAL4–AD1 Figure 1. BRCA1 interacts with JunB and JunD in yeast and was coexpressed with JunB or JunD that was fused to a mammalian two-hybrid assays. (A) Summary of the results from potent transcriptional activation domain (VP16). Consis- the yeast two-hybrid screen. The plus sign (+) indicates rapid tent with previous findings (Hu et al. 2000), GAL4–AD1 growth of the yeast cells in the selective medium as a result of alone did not significantly activate transcription in elevated expression of the HIS3 gene. Also presented is a sche- matic diagram of the BRCA1 protein, illustrating the location of HEK293T cells (Fig. 1B, cf. column 1 with 4), whereas the trans-activation domains. The solid bar within AD1 indi- GAL4–AD2 functioned as a potent activator in the same cates the coiled-coil motif. (B) Mammalian two-hybrid assay cellular context (Fig. 1B, column 7). Coexpression of showing the interactions between BRCA1 and the Jun proteins. VP16–JunB resulted in a significant elevation of tran- HEK293T cells were transfected with the mammalian bait and scriptional activation by GAL4–AD1 (Fig. 1B, cf. column prey constructs. Shown in the y axis is the fold of increase in the 4 with 5). VP16–JunD also had a similar, albeit less pro- luciferase activity over the negative control (column 1; GAL4– nounced, effect on GAL4–AD1 (Fig. 1B, column 6). In DBD and VP16). contrast, the same prey constructs did not enhance the activity of GAL4–DBD (Fig. 1B, columns 1–3) or GAL4– examined. Thus, the physical and functional link be- AD2 (Fig. 1B, columns 7–9). Expression of the GAL4 de- tween BRCA1 and JunB may be important for their func- rivatives was unaffected by the VP16–Jun proteins (data tions in tissue-specific transcriptional regulation and not shown). Thus, the results from both yeast and mam- suppression of tumor development. malian two-hybrid systems indicate that JunB and JunD interact with AD1 in vivo. Results JunB and JunD interact with AD1 of BRCA1 BRCA1 specifically interacts with the Jun proteins in the yeast and mammalian two-hybrid systems of the AP1 family To identify the potential partner(s) of BRCA1 that medi- To verify the two-hybrid findings, we examined the abil- ate AD1 function in transcriptional activation, we con- ity of various AP1 family members to interact with the

1510 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

BRCA1–Jun interaction native full-length BRCA1 in human cells. FLAG-tagged c-Jun, JunB, JunD, and c-Fos were expressed in HEK293T or ES2 cells (Fig. 2A, top). Following immunoprecipitation with an anti-FLAG antibody, the presence of the endogenous BRCA1 in the immunoprecipitates was detected by immunoblotting with an anti-BRCA1 antibody (Fig. 2A). Consistent with the two-hybrid results, the FLAG-tagged JunB and JunD were associated with native BRCA1 (Fig. 2A, lanes 2 and 4). In addition, BRCA1 was also coprecipitated with the FLAG-tagged c-Jun (Fig. 2A, lane 3). The in vivo association of BRCA1 and the Jun proteins was unlikely to be mediated by nucleic acids, as it was not affected by the treatment of nuclease or ethid- ium bromide (data not shown). In contrast to the Jun proteins, the FLAG-tagged c-Fos did not bind to native BRCA1 in either HEK293T (Fig. 2A, lane 5) or ES2 cells (Fig. 2A, lane 9). For reasons that will become obvious later, the co-IP experiment in HEK293T cells was also repeated in the presence of ectopically expressed HA-JunB. Once again, no endogenous BRCA1 was detected in the FLAG-cFos immunoprecipi- tate (Fig. 2A, lane 6). It is known that c-Fos forms heterodimers with the Jun proteins, but not homodimers with itself in vivo (Karin et al. 1997). Endogenous Jun proteins were coimmunoprecipitated with the FLAG- cFos (data not shown). Thus, our finding suggests that BRCA1 does not bind to either c-Fos monomers or Jun- Fos heterodimers. Next, we sought to ascertain the interaction between BRCA1 and the Jun proteins in a more direct manner. The bZIP region of various AP1 proteins was fused with glutathione S-transferase (GST). The purified GST proteins were immobilized on glutathione beads and incubated with the in vitro translated, 35S-labeled AD frag- Figure 2. Interaction of the Jun proteins with BRCA1. (A) Hu- ment of BRCA1. As shown in Figure 2B, AD was pulled man HEK293T (lanes 1–6) or ES2 (lanes 7–9) cells were transfected with the various expression vectors for the FLAG-tagged down by all three GST–Jun fusion proteins (Fig. 2B, lanes AP-1 proteins. Following immunoprecipitation (IP) with an 2–4), but not by GST alone (Fig. 2B, lane 1) or GST–cFos anti-FLAG antibody, the proteins in the immunoprecipitates (Fig. 2B, lane 5). This result suggests a direct interaction were resolved by SDS-PAGE, and the presence of the endog- between BRCA1 and the Jun proteins. Furthermore, the enous BRCA1 proteins was detected by immunoblotting (WB) in vitro finding confirms that the bZIP region of the Jun with an anti-BRCA1 antibody (bottom). Equal amounts of the proteins is sufficient for binding to BRCA1. cell lysates were analyzed by immunoblotting for the expres- To ascertain the BRCA1–Jun interaction in a more sion of the FLAG-tagged AP-1 proteins (top). Lane 6 shows that physiological context, the endogenous AP1 proteins c-Fos does not interact with BRCA1, even in the presence of 35 from ES2 cells was immunoprecipitated with var- exogenous HA-tagged JunB. (B) S-labeled BRCA1–AD was ious commercially available antibodies. Subsequent im- made with an in vitro translation kit (Promega), and incubated with the GST alone (lane 1) or GST fused with the bZIP region munoblotting with an anti-BRCA1 antibody showed of various AP1 proteins (lanes 2–5) that were immobilized on that the endogenous BRCA1 was coprecipitated with glutathione beads. After extensive washing, the coprecipitated both c-Jun and JunB, but not c-Fos (Fig. 2C, cf. lanes 2, 4, proteins were analyzed by SDS-PAGE and fluorography. (C) Co- and 6–8). In addition, the BRCA1 signals were dimin- immunoprecipitation of native BRCA1 and native Jun proteins ished when two antibody-specific competing peptides in ES2 cells. Lysates of ES2 cells were immunoprecipitated with were included in the immunoprecipitation reactions different anti-AP1 antibodies (sc-44 for pan-Jun, sc-45 and sc- (Fig. 2C, cf. lane 2 with 3 and 4 with 5). A reciprocal 1694 for c-Jun, sc-8051 for JunB, and sc-52 for c-Fos). The pres- co-IP experiment using two different anti-BRCA1 anti- ence of BRCA1 in the immunoprecipitates was detected by bodies also shows the physical association between Western blotting using an anti-BRCA1 antibody (Ab-1 from On- BRCA1 and c-Jun. Interestingly, the BRCA1–Jun inter- cogene). In lanes 3 and 5, an excess of the corresponding competing peptides was included in the immunoprecipitation reac- action is refractory to a fairly high-salt and deter- tions. (D) A reciprocal co-IP was performed in ES2 cell lysates gent concentration (500 mM NaCl and 1% NP-40). using either anti-␣-tubulin (as a negative control) or anti- Taken together, the results strongly indicate an in vivo BRCA1 antibodies (Ab1 and Ab3; Oncogene) in the immuno- association of BRCA1 with specific members of the AP1 precipitation. The blot was probed with an anti-cJun antibody family. (sc-1694).

GENES & DEVELOPMENT 1511 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Hu and Li

AD1 is necessary and sufficient for BRCA1 binding to the Jun proteins To further characterize the Jun-binding domain in BRCA1, we constructed a series of deletional mutants and analyzed their ability to interact with the FLAG– JunB using the coimmunoprecipitation assay. The majority of cancer-predisposing mutations in BRCA1 result in truncation of the protein. As summarized in Figure 3, a disease-associated BRCA1 truncation mutant that retained the coiled-coil region in AD1 still interacted with BRCA1 (Fig. 3, construct 2), whereas those that lacked the AD1 region failed to bind to JunB (Fig. 3, constructs 3 and 4). In a different set of mutational analyses, multiple fragments that spanned the entire BRCA1 protein were tested for JunB binding (Fig. 3, constructs 5–12). The only fragments that displayed significant affinity for JunB were those that contained AD1 (Fig. 3, constructs 10 and 12). Further deletional studies within AD1 (Fig. 3, constructs 13–16) revealed a minimal JunB-binding domain (Fig. 3, construct 15; amino acids 1343–1440), which consists of the coiled-coil motif and an 60 amino acid upstream sequence. This finding indicates that the AD1 region is both necessary and sufficient for BRCA1 binding to JunB.

The coiled-coil motif in BRCA1 is critical for binding to Jun and for AD1-mediated transcriptional activation To establish a stronger link between the BRCA1–Jun interaction and AD1-dependent transcriptional activation,

Figure 4. BRCA1-Jun interaction correlates with AD1-mediated transcriptional activation. (A) HEK293T cells were cotransfected with the expression vectors for the FLAG-tagged JunB and HA-tagged GAL4–AD domain. Lysates from the transfected cells were immunoprecipitated (IP) with an anti-FLAG antibody, and the immunoprecipitates were probed by immunoblotting (WB) with an anti-HA antibody (bottom). As a control, expression of the wild-type and mutant GAL4–AD fusion proteins was detected by immunoblotting of the crude lysates using the anti-HA antibody (top). (B) HA-tagged full-length BRCA1 proteins were coexpressed with FLAG–JunB in HEK293T cells. Anti-FLAG immunoprecipitation was followed by immunoblotting using the anti-HA antibody. (C) ES-2 cells were cotransfected with the GAL4–AD expression vectors and a GAL4-responsive luciferase reporter plasmid. Also shown at top of the graph is a Western blot for the wild-type and mutant fusion proteins. The relative transcriptional activity of the wild- type GAL4–AD construct is set at 100.

we introduced various point mutations into the coiled- coil region of AD1 (Fig. 4). All mutants were expressed at similar levels as the wild-type proteins (Fig. 4A,C). The Figure 3. Characterization of the Jun-binding domain in mutational effect on the BRCA1–JunB interaction was BRCA1. Various HA-tagged BRCA1 fragments were ectopically tested in the contexts of both AD (Fig. 4A) and the full- coexpressed with FLAG–JunB in HEK293T cells. Immunopre- cipitation was carried out using an anti-FLAG antibody. The plus length BRCA1 (Fig. 4B). Extensive work on other coiled- sign indicates a significant signal of the BRCA1 fragments in the coil proteins has shown that the leucine residues at po- anti-FLAG immunoprecipitates. The hatched box indicates AD1, sition d of the heptad repeat are the critical determinants whereas the shaded boxes designate the two BRCT repeats in for the coiled-coil structure (Lupas et al. 1991; Lupas AD2. The solid bar within AD1 represents the coiled-coil region. 1996). Substitution of two such leucines in the coiled-

1512 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

BRCA1–Jun interaction coil region of BRCA1 (L1407 and L1414) abrogated the able level in most of the cell lines examined (Fig. 5A), it BRCA1–JunB interaction (Fig. 4A,B, lanes 3–5). Impor- could still be limiting for supporting AD1-mediated tran- tantly, the same mutations also impaired the transcrip- scriptional activation. For example, despite the higher tional activity of GAL4–AD in the luciferase reporter JunB expression in ES2 cells than that in HEK293T cells, assay (Fig. 4C, lanes 3–5). In contrast, mutations that AD1 alone only moderately activates transcription in presumably did not affect the coiled-coil structure ES2 cells (Hu et al. 2000) (Fig. 5C, cf. lanes 1 and 6). (I1405V, H1402Y, and H1421Y) still retained the affinity Ectopic expression of JunB, and JunD to a lesser extent, of BRCA1 for JunB (Fig. 4A,B, lanes 6–8). Likewise, the enhanced AD1-mediated transcriptional activation in GAL4 fusion proteins carrying these mutations were ES2 cells (Fig. 5C, columns 7 and 9). As observed in still capable of stimulating transcription (Fig. 4C, lanes HEK293T cells, c-Jun and c-Fos failed to confer such su- 6–8). All of the mutations shown in Figure 4 had the peractivation of AD1 (Fig. 5C, columns 8and 10). There- same effect on c-Jun and JunD binding as they did on fore, our data indicate that the strength of AD1 in tran- JunB (data not shown). Thus, the affinity of BRCA1 for scriptional activation is limited by the cellular level of the Jun proteins strongly correlates with the strength of JunB. An interesting difference between HEK293T and the trans-activation domain in transcriptional stimula- ES2 cells with regard to AD1 function is that ES2, but tion. not HEK293T cells, can support a synergistic action of AD1 and AD2 (Hu et al. 2000). It is therefore plausible that the level of JunB in ES2 cells, although insufficient JunB potentiates AD1 function in transcriptional for supporting maximal activation by AD1 alone, may be activation replete for the cooperative activation by AD1 and AD2. Previous characterization of AD1 indicates that this Given the proximity of the Jun-binding domain to the trans-activation domain functions in a cell context-de- previously defined transition point of BRCA1 mutations pendent manner and that it displays a less robust tran- that are associated with higher ovarian cancer risk (Gay- scription activity than AD2. For example, HEK293T ther et al. 1995), we speculated that this domain likely cells were deficient in AD1-mediated transcriptional ac- provides special protective functions against ovarian tivation, despite their ability to support high-transfec- cancer. In such an event, a deficit of JunB in ovarian tion efficiency and robust expression of the GAL4 deriva- epithelium might contribute to development of ovarian tives (Hu et al. 2000). Given the specific interactions cancer in particular. To test this possibility, we com- between AD1 and the Jun proteins, we speculated that pared the JunB mRNA level in tumor and normal tissues the lack of AD1 function in HEK293T cells might be due from the same individuals. The normalized real-time to limited expression of one or more Jun proteins. Im- PCR results are shown in Figure 5D. In seven of the nine munoblotting of crude lysates from HEK293T cells indi- matched cDNA pairs from ovary tissue, JunB expression cated that the protein level of JunB was extremely low is significantly lower in tumor than in the normal issues (Fig. 5A, lane 2), as has been observed by others (Bakiri et (Fig. 5D, pairs 1–5, 7, and 8). The remaining two ovarian al. 2000). In contrast, the levels for c-Jun, JunD, and c-Fos pairs (Fig. 5D, pairs 6 and 9) had very low JunB mRNA in HEK293T cells were comparable with those in the levels even in the normal tissues, suggesting that there other cell lines examined (Fig. 5A). might be some intrinsic abnormality in these two nor- To determine whether the low level of JunB protein in mal cases. Interestingly, the differential expression of HEK293T cells was causally related to the lack of AD1 JunB in most ovarian pairs was not obvious in a panel of activity, we asked whether the deficiency in supporting matched cDNA pairs from breast tissue (Fig. 5D, pairs AD1 function could be rescued by ectopic expression of 10–16). This finding is consistent with the notion that JunB. As shown in Figure 5B, coexpression of FLAG– the BRCA1–JunB interaction may play a role in specific JunB and GAL4–AD1 significantly enhanced the ability suppression of ovarian cancer development. of AD1 to activate transcription (Fig. 5B, cf. lanes 6 and 7). However, JunB did not superactivate GAL4–AD2 (Fig. Discussion 5B, cf. lanes 11 and 12), nor did it rescue the transcriptional defect of a coiled-coil mutant of AD1 that failed to A wealth of evidence strongly suggests that BRCA1 plays bind JunB (L1407P; Fig. 5B, cf. lanes 16 and 17). These an important role in the maintenance of genome stabil- results strongly indicate that JunB potentiates BRCA1 ity via its function in transcriptional regulation and DNA function through its interaction with the coiled-coil re- repair. However, it remains puzzling that disease-associ- gion of AD1. In contrast to JunB, ectopic expression of ated mutations in BRCA1, which compromise such uni- c-Jun, JunD, or c-Fos failed to complement the deficiency versal nuclear functions as transcription and DNA repair, of HEK293T cells (Fig. 5B, lanes 8–10), despite their expres- specifically lead to elevation of the risk in developing sion levels equivalent to that of junB (Fig. 2A). This differ- breast and ovarian cancers. It has been suggested that the ential effect of the AP1 proteins was observed at multiple rapid proliferating status of the breast epithelium during concentrations of the AP1 expression vectors (data not puberty could render it particularly susceptible to BRCA1 shown). Thus, although all three Jun proteins are capable of mutation-dependent tumorigenesis (Scully and Livingston binding to BRCA1, JunB exhibits a distinct function in fa- 2000). In addition, the fact that breast and ovary are both cilitating AD1-mediated transcriptional activation. estrogen-responsive tissues and that BRCA1 can modulate Although JunB protein was expressed above the detect- transcriptional activation by estrogen receptor could also

GENES & DEVELOPMENT 1513 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Hu and Li

Figure 5. JunB can superactivate AD1-dependent transcriptional activation. (A) An equal amount of cell lysates from various cell lines was probed by immunoblotting with specific antibodies against c-Jun, JunB, JunD, and c-Fos. (B) HEK293T cells were cotransfected with the luciferase reporter plasmid, the GAL4 derivatives, and various AP1 expression vectors. The relative luciferase activity in the presence of GAL4–DBD alone (column 1) is set at one. (C) Luciferase reporter assay was performed in ES2 cells in the same manner as described in B.(D) Normalized matched cDNA pairs of normal and tumor ovarian (1–9) or breast (10–16) tissues were analyzed by the real-time PCR reactions for JunB mRNA expression.

explain the organ-specific nature of the BRCA1-dependent of one of the trans-activation domains of BRCA1 (AD1). cancer risk (Fan et al. 1999, 2001; Zheng et al. 2001). It is First, both in vitro and in vivo experiments indicate that also possible that the tissue-specific action of BRCA1 may the two proteins interact through the coiled-coil domain be determined by more than one BRCA1-associated pro- of BRCA1 and the bZIP domain in JunB. The affinity of tein complex. In such an event, changes in the level and/or BRCA1 for JunB is strongly correlated with the strength biochemical properties of a number of the BRCA1-associ- of the AD1 domain in transcriptional activation. Fur- ated proteins could contribute to the development of neo- thermore, the data suggest that the cellular level of the plasm in these tissues. JunB protein is an important determinant for AD1 func- The findings in the current study strongly suggest that tion in transcriptional activation. Limited AD1 tran- JunB plays an important role in mediating the function scriptional activity due to a deficit of JunB can be rescued

1514 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

BRCA1–Jun interaction by ectopic expression of JunB. This finding provides a (Bakiri et al. 2000), and at the same time activate the molecular explanation for the previously observed cell- transcription of p16INK4a, an inhibitor of the cyclin type dependent behavior of AD1 (Hu et al. 2000). Lastly, D-CDK4–D-CDK6 complexes (Passegue and Wagner JunB mRNA levels are down-regulated in many ovarian 2000). Recent work with conditional JunB knockout tumor tissues compared with the normal tissues from mice also supports a role of JunB in tumor suppression the same individuals. Given that the AD1 domain of (Passegue et al. 2001). Thus, although our study does not BRCA1 encompasses the demarcation point for BRCA1 exclude a potential functional link between BRCA1 and mutations that are associated with increased risk of the other two Jun proteins, the cooperation between ovarian cancer, our results raise an intriguing possibility BRCA1 and JunB in transcriptional regulation may be that the BRCA1–JunB interaction may be particularly related to their known functions in suppression of tis- important for suppression of ovarian cancer. However, sue-specific tumor development. due to the lack of detailed information concerning the clinical characteristics of the patients and the exact cell types from which the cDNA pairs were prepared, further Materials and methods study is needed to explore the exact biological signifi- cance of the differential JunB expression observed in the Yeast two-hybrid screen current work. The standard yeast two-hybrid screen was performed in the fol- The exact geometry and stoichiometry of the BRCA1– lowing manner. First, bait plasmid was generated by inserting a Jun complexes remain to be determined. It is also pos- PCR-amplified cDNA fragment encoding AD1 (amino acids sible that the BRCA1–Jun interaction is aided by addi- 1293–1559) of BRCA1 into pGBT8(Clontech), resulting in an tional factors. The finding that BRCA1 is not associated in-frame fusion with the GAL4 DNA-binding domain (DBD). with the cFos-containing complexes strongly suggests Second, the resultant plasmid, pGBT8-AD1, and a human ovary that Jun-Fos dimers, which is the predominant heterodi- cDNA library (Clontech) were cotransformed into the Saccha- meric form of the AP1 family in vivo, are not capable of romyces cerevisiae reporter strain Hf7C according to the manu- binding to BRCA1. This raises the possibility that facturer’s instructions (Clontech). Transformants were plated on synthetic medium lacking tryptophan, leucine, and histi- BRCA1 may specifically target Jun monomers and/or dine, but containing 100 mM 3-aminotriazole (3-AT), which Jun–Jun dimers. Conceivably, this Jun-specific interac- was used to suppress the relatively high background due to the tion may lead to changes in the relative abundance, sub- intrinsic transcripional activity of the bait construct. Approxi- nuclear localization, and biochemical characteristics of mately 26 million transformants were screened, of which 32 various forms of the AP1 proteins. The bZIP sequences were judged to be strongly HIS-positive. Additional screens us- among the Jun proteins are highly conserved, whereas ing a number of negative controls were carried out (see Fig. 1A), those of Fos and Jun are relatively divergent. This could and only those candidates exhibiting AD1-specific interaction explain the disparity in their binding affinity for BRCA1. were characterized further. Twenty clones contained partial It is tempting to speculate that the leucine zipper motif cDNA sequences for JunD, and six contained the JunB cDNA may only provide the architectural basis for binding to sequences. BRCA1. Additional amino acid residues in the bZIP region that are unique to the Jun proteins may serve as the Mammalian cell transfection and luciferase assay actual contact points for the coiled-coil region of HEK293T cells were maintained in DMEM medium with 10% BRCA1. fetal calf serum and were transfected using LipofectAmine 2000 Whereas all three Jun proteins are capable of interact- (GIBCO BRL). ES-2 cells were grown in McCoy’s 5A medium ing with BRCA1, only JunB exhibits a strong enhancing with 10% fetal calf serum and were transfected using Lipofect- effect on the AD1 transcriptional activity. Such a func- Amine Plus (GIBCO BRL). In a typical GAL4-based transcriptional difference among the Jun proteins could be attrib- tion reporter assay performed in HEK293T cells, 0.5 µg reporter uted to the more divergent sequences outside of the bZIP and 1.0 µg protein expression vectors were used. Half of these domain. The observation that BRCA1 can selectively amounts were used for the luciferase assays in ES2 cells. The luciferase assays were performed as described previously (Hu et target specific members of the AP1 family for physical al. 2000). The expression vectors for GAL4 derivatives used in and functional interaction may have profound biological the mammalian two-hybrid assay were described previously ramifications. AP-1 family members form a large num- (Hu et al. 2000). The vectors for the VP16 fusion proteins used ber of homodimers and heterodimers in vivo, each of in the mammalian two-hybrid assay were constructed by clon- which may exhibit distinct regulatory properties (Karin ing the cDNA sequences for JunB and JunD into the prey plas- et al. 1997). As a consequence, different AP1 family mid as described previously (Yu et al. 1998). members can play diverse and even opposing roles in cell proliferation and differentiation. For example, it has Coimmunoprecipitation been well documented that c-Jun is positively involved The FLAG-tagged AP1 proteins were expressed from the in cell proliferation and Ras-mediated oncogenesis pcDNA3 vector (Clontech). Plasmids pCG-HA-GAL4(1–94)–AD (Mechta-Grigoriou et al. 2001; Shaulian and Karin 2001). (Hu et al. 2000) and pcDNA3␤-HA-BRCA1 (full length) (Scully On the other hand, JunB and JunD can suppress Ras- and et al. 1997) were described previously. Twenty-four hours after Src-induced cellular transformation (Johnson et al. 1996; transfection, cells were washed twice with PBS and lysed in 0.5 Mechta et al. 1997). Moreover, JunB can antagonize the mL Lysis Buffer (50 mM HEPES at pH 8.0, 250 mM NaCl, stimulatory function of c-Jun in cyclin D1 transcription 0.1%NP-40, and protease inhibitor tablets from Roche-

GENES & DEVELOPMENT 1515 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

Hu and Li

Bohringer). After brief sonication, the lysate was centrifuged at tions for primers and probe were 300 and 250 nM, respectively. 14,000 rpm for 12 min at 4°C. The supernatant was used for cDNA pairs (Clontech) normalized with ␤-actin and/or ribo- subsequent coimmunoprecipitation. Fifteen microliters of 50% somal protein S9 were used as the templates in the PCR reac- slurry of the anti-FLAG agarose beads (Sigma) were used in each tions. In Figure 5D, pairs 1–4 come from HP101O, HP103O, immunoprecipitation, and the tubes were rotated overnight at HP104O, and HP105O; pairs 5–9 from a human ovary-matched 4°C. Following incubation, the beads were centrifuged at 3300 cDNA pair panel (K1435-1); pairs 10–11 from HB102B and rpm for 2 min, and washed once with the lysis buffer and twice HP104B; and pairs 12–16 from a human breast-matched cDNA with a washing buffer (50 mM HEPES at pH 8.0, 500 mM NaCl, pair panel (K1432-1). 1% NP-40), with each wash lasting at least 1 h. The precipitates were then eluted in 10 µL 2× protein sample buffer and loaded on SDS–polyacrylamide, followed by Western blotting accord- Acknowledgments ing to the standard procedures. Five microliters of the input crude extract were used for detecting protein expression levels. We thank Richard Baer and Ralph Scully for sharing reagents; The presence of the endogenous BRCA1 was detected using a Qinong Ye, Hongjun Zhong, and Anne Allison for technical commercially available anti-BRCA1 antibody (Ab-1 from Onco- assistance; Mark Alexandrow, Christian Loch, Sarah Aiyar, and gene). The HA-tagged proteins were detected using an anti-HA Sagar Ghosh for critical reading of the manuscript; and mem- monoclonal antibody (12CA5). bers of the Li laboratory for stimulating discussion. This work The coimmunoprecipitation experiment shown in Figure 2C was supported by grants from the National Cancer Institute and was conducted using lysates from ES2 cells, anti-Jun antibodies Department of Defense. for immunoprecipitation, and an anti-BRCA1 antibody (Ab-1 The publication costs of this article were defrayed in part by from Oncogene) for immunoblotting. A total of 1.5 µg of the payment of page charges. This article must therefore be hereby following commercially available anti-AP1 antibodies (Santa marked “advertisement” in accordance with 18USC section Cruz Biotech.) were used in immunoprecipitation: sc-44 (anti- 1734 solely to indicate this fact. cJun, JunB, and JunD); sc-45 (anti-cJun); sc-1694 (anti-cJun); sc- 8051 (anti-JunB); sc-52 (anti-cFos). Two competing peptides for c-Jun in immunoprecipitation (sc-44p and sc-52p; Santa Cruz References Biotech.) were used. The reciprocal co-IP shown in Figure 2D was done using 2.5 µg of anti-BRCA1 (Ab1 and Ab3) or anti-␣- Bakiri, L., Lallemand, D., Bossy-Wetzel, E., and Yaniv, M. 2000. tubulin antibodies (Ab-1; Oncogene) for IP and an anti-cJun an- Cell-cycle-dependent variations in c-Jun and JunB phos- tibody (sc-1694) for immunoblotting. phorylation: A role in the control of cyclin D1 expression. EMBO J. 19: 2056–2068. Bochar, D.A., Wang, L., Beniya, H., Kinev, A., Xue, Y., Lane, In vitro GST pulldown assay W.S., Wang, W., Kashanchi, F., and Shiekhattar, R. 2000. The PCR fragments encoding the bZIP region of the AP1 pro- BRCA1 is associated with a human SWI/SNF-related com- teins were fused in-frame with the GST portion in plasmid plex: Linking chromatin remodeling to breast cancer. Cell pGEX-2T (Pharmacia). The GST–bZIP proteins were expressed 102: 257–265. and purified according to the manufacturer’s instruction, with Chapman, M.S. and Verma, I.M. 1996. Transcriptional activa- the induction of the protein expression performed at 37°C for 3 tion by BRCA1. Nature 382: 678–679. h. The fusion gene encoding FLAG-tagged BRCA1–AD in the Chinenov, Y. and Kerppola, T.K. 2001. Close encounters of pcDNA3 vector was also under the control of the bacteriophage many kinds: Fos-Jun interactions that mediate transcription T7 promoter. This plasmid was used for in vitro transcription regulatory specificity. Oncogene 20: 2438–2452. and translation in the TnT Reticulocyte Lysate system (Pro- Fan, S., Wang, J., Yuan, R., Ma, Y., Meng, Q., Erdos, M.R., Pes- mega). The 35S-labeled FLAG–BRCA1–AD was mixed with 10 tell, R.G., Yuan, F., Auborn, K.J., Goldberg, I.D., et al. 1999. µg of GST derivatives bound to agarose beads in 0.5 mL of bind- BRCA1 inhibition of estrogen receptor signaling in trans- ing buffer (50 mM HEPES at pH 8.0, 150 mM NaCl, 1 mM fected cells. Science 284: 1354–1356. EDTA, 0.1% NP-40 and protease inhibitor tablets). The binding Fan, S., Ma, Y.X., Wang, C., Yuan, R.Q., Meng, Q., Wang, J., reaction was carried out at 4°C overnight and subsequently Erdos, A.M., Goldberg, I.D., Webb, P., Kushner, P.J., et al. washed four times with the washing buffer (the same as binding 2001. Role of direct interaction in BRCA1 inhibition of es- buffer, except 0.5% NP-40), 30 min each round. The beads were trogen receptor activity. Oncogene 20: 77–87. eluted in 10 µL of 2× protein sample buffer and the proteins Gayther, S.A., Warren, W., Mazoyer, S., Russell, P.A., Harring- were resolved on 10% SDS–polyacrylamide. The gel was then ton, P.A., Chiano, M., Seal, S., Hamoudi, R., van Rensburg, dried and exposed to X-ray films overnight. E.J., Dunning, A.M., et al. 1995. Germline mutations of the BRCA1 gene in breast and ovarian cancer families provide evidence for a genotype-phenotype correlation. Nature Real-time PCR Genet. 11: 428–433. The ABI PRISM 7700 Sequence Detection System (PE Biosys- Harkin, D.P., Bean, J.M., Miklos, D., Song, Y.-H., Truong, V.B., tems) was used for the quantitative analysis of JunB mRNA Englert, C., Christians, F.C., Ellisen, L.W., Maheswaran, S., expression. The following primers were designed using the Oliner, J.D., et al. 1999. Induction of GADD45 and JNK/ Primer Express software provided by the manufacturer. JunB SAPK-dependent apoptosis following inducible expression of forward primer, ACTCATACACAGCTACGGGATACG; JunB BRCA1. Cell 97: 575–586. reverse primer, GGCTCGGTTTCAGGAGTTTG; TagMan probe, Hu, Y.F., Hao, Z.L., and Li, R. 1999. Chromatin remodeling and VICCCCCTGGTGGCCTCTCTCTACACGATAMRA. activation of chromosomal DNA replication by an acidic Primers, probes, and TagMan universal PCR master mixes transcriptional activation domain from BRCA1. Genes & were purchased from Applied Biosystems. PCR was performed Dev. 13: 637–642. in a 50-µL reaction volume, using the default assay condition in Hu, Y.-F., Miyake, T., Ye, Q., and Li, R. 2000. Characterization the ABI PRISM 7700 Sequence Detector. The final concentra- of a novel trans-activation domain of BRCA1 that functions

1516 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

BRCA1–Jun interaction

in concert with the BRCA1 C-terminal (BRCT) domain. J. ponents of the histone deacetylase complex. Proc. Natl. Biol. Chem. 275: 40910–40915. Acad. Sci. 96: 4983–4988. Johnson, R., Spiegelman, B., Hanahan, D., and Wisdom, R. 1996. Yu, X., Wu, L.C., Bowcock, A.M., Aronheim, A., and Baer, R. Cellular transformation and malignancy induced by ras re- 1998. The C-terminal (BRCT) domains of BRCA1 interact in quire c-Jun. Mol. Cell. Biol. 16: 4504–4511. vivo with CtIP, a protein implicated in the CtBP pathway of Karin, M., Liu, Z., and Zandi, E. 1997. AP-1 function and regu- transcriptional repression. J. Biol. Chem. 273: 25388–25392. lation. Curr. Opin. Cell Biol. 9: 240–246. Zhang, H., Somasundaram, K., Peng, Y., Tian, H., Zhang, H., Bi, Lupas, A. 1996. Coiled coils: New structures and new functions. D., Weber, B.L., and El-Reiry, W.S. 1998a. BRCA1 physically Trends Biosci. 21: 375–382. associated with p53 and stimulates its transcriptional activ- Lupas, A., Van Dyke, M., and, Stock, J. 1991. Predicting coiled ity. Oncogene 16: 1713–1721. coils from protein sequences. Science 252: 1162–1164. Zhang, H., Tombline, G., and Weber, B.L. 1998b. BRCA1, MacLachlan, T.K., Somasundaram, K., Sgagias, M., Shifman, Y., BRCA2, and DNA damage response: Collision or collusion? Muschel, R.J., Cowan, K.H., and El-Deiry, W.S. 2000. Cell 92: 433–436. BRCA1 effects on the cell cycle and DNA damage response Zheng, L., Li, S., Boyer, T.G., and Lee, W.H. 2000. Lessons are linked to altered gene expression. J. Biol. Chem. learned from BRCA1 and BRCA2. Oncogene 19: 6159–6175. 275: 2777–2785. Zheng, L., Annab, L.A., Afshari, C.A., Lee, W.-H., and Boyer, Mechta, F., Lallemand, D., Pfarr, C.M., and Yaniv, M. 1997. T.G. 2001. BRCA1 mediates ligand-independent transcrip- Transformation by ras modifies AP-1 composition and activ- tional repression of the estrogen receptor. Proc. Natl. Acad. ity. Oncogene 14: 837–847. Sci. 98: 9587–9592. Mechta-Grigoriou, F., Gerald, D., and Yaniv, M. 2001. The mammalian Jun proteins: Redundancy and specificity. On- cogene 20: 2378–2389. Monteiro, A.N.A. 2000. BRCA1: Exploring the links to transcription. Trends Biosci. 25: 469–474. Monteiro, A.N., August, A., and Hanafusa, H. 1996. Evidence for a transcriptional activation function of BRCA1 C-terminal region. Proc. Natl. Acad. Sci. 93: 13595–13599. Neish, A.S., Anderson, S.F., Schlegel, B.P., Wei, W., and Parvin, J.D. 1998. Factors associated with the mammalian RNA polymerase II holoenzyme. Nucleic Acids Res. 26: 847–853. Ouchi, T., Monteiro, A.N., August, A., Aaronson, S.A., and Hanafusa, H. 1998. BRCA1 regulates p53-dependent gene expression. Proc. Natl. Acad. Sci. 95: 2302–2306. Pao, G.M., Janknecht, R., Ruffner, H., Hunter, T., and Verma, I.M. 2000. CBP/p300 interact with and function as transcriptional coactivators of BRCA1. Proc. Natl. Acad. Sci. 97: 1020–1025. Parvin, J.P. 2001. BRCA1 at a branch point. Proc. Natl. Acad. Sci. 98: 5952–5954. Passegue, E. and Wagner, E.F. 2000. JunB suppresses cell proliferation by transcriptional activation of p16(INK4a) expression. EMBO J. 19: 1969–1979. Passegue, E., Jochum, W., Schorpp-Kistner, M., Mohle-Stein- lein, U., and Wagner, E.F. 2001. Chronic myeloid leukemia with increased granulocyte progenitors in mice lacking JunB expression in the myeloid lineage. Cell 104: 21–32. Rahman, N. and Stratton, M.R. 1998. The genetics of breast cancer susceptibility. Annu. Rev. Genet. 32: 95–121. Scully, R. and Livingston, D.M. 2000. In search of the tumor- suppressor functions of BRCA1 and BRCA2. Nature 408: 429–432. Scully, R., Anderson, S.F., Chao, D.M., Wei, W., Ye, L., Young, R.A., Livingston, D.M., and Parvin, J.D. 1997. BRCA1 is a component of the RNA polymerase II holoenzyme. Proc. Natl. Acad. Sci. 94: 5605–5610. Shaulian, E. and Karin, M. 2001. AP-1 in cell proliferation and survival. Oncogene 20: 2390–2400. Somasundaram, K., Zhang, H., Zeng, Y.X., Houvras, Y., Peng, Y., Wu, G.S., Licht, J.D., Weber, B.L., and El-Deiry, W.S. 1997. Arrest of the cell cycle by the tumour-suppressor BRCA1 requires the CDK-inhibitor p21WAF1/CiP1. Nature 389: 187–190. Welcsh, P.L. and King, M.-C. 2001. BRCA1 and BRCA2 and the genetics of breast and ovarian cancer. Hum. Mol. Genet. 10: 705–713. Yarden, R.I. and Brody, L.C. 1999. BRCA1 interacts with com-

GENES & DEVELOPMENT 1517 Downloaded from genesdev.cshlp.org on September 28, 2021 - Published by Cold Spring Harbor Laboratory Press

JunB potentiates function of BRCA1 activation domain 1 (AD1) through a coiled-coil-mediated interaction

Yan-Fen Hu and Rong Li

Genes Dev. 2002, 16: Access the most recent version at doi:10.1101/gad.995502

References This article cites 36 articles, 15 of which can be accessed free at: http://genesdev.cshlp.org/content/16/12/1509.full.html#ref-list-1

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.

Cold Spring Harbor Laboratory Press